单通道通信信号盲分离算法

Blind Separation Algorithm of Single Channel Communication Signals

针对空间侦察中单通道宽带接收机截获到多个独立辐射源信号时的盲源分离问题,提出了一种基于希尔伯特黄变换(HHT)的单通道通信信号源数估计,以及采用独立分量分析(ICA)实现盲源分离的方法。上述算法首先对单通道混合通信信号进行希尔伯特黄(HHT)变换,得到混合信号的边际谱。然后通过边际谱得到通信信号源数目,根据源信号数目以及本征模态函数(IMF)重组多通道混合通信信号,并通过独立分量分析(ICA)算法来恢复原始信号。仿真结果表明,分离前后信号的相似系数大于0.9,改进算法能够有效地估计单通道通信信号源数目并且准确分离盲源。

Contrary to the blind source separation of a plurality of independent radiation source singles intercepted by single-channel wideband receiver for space reconnaissance, a novel method was proposed, which is based on Hil- bert-Huang Transform（HHT） to estimate the number of the single channel communication signals, and Independent Component Analysis（ICA） to separate the blind source signals. Firstly, the HHT was applied on the single channel mixed communication signals and the marginal spectrum of this mixed signals was obtained. Then we estimated the source number of the communication signals with the marginal spectrum, reorganized the muhichannel mixed commu- nication signals according to the source signal number and its Intrinsic Mode Function （ IMF）, and recovered the source signals by the method of ICA. The results of simulation experiments show that the similarity coefficient of the signals separation is larger than 0.9 and the proposed algorithm can estimate the number of the single channel commu- nication signals effectively and separate the blind source accurately.